Computation of molecular electronic structure by genetic algorithm

Abstract

Several strategies based on Genetic Algorithms have been explored to locate the ground state energy and structure of atoms and molecules. A variational recipe in a finite basis has been invoked in one of the strategies leading to a matrix eigenvalue problem that has been solved by GA with simultaneous optimization of the basis. In the second approach, the molecular electronic Schrodinger equation has been solved by Genetic Algorithm by directly optimizing the electronic probability amplitude distribution in space. In the third approach, the single-particle density matrix in atomic orbital basis has been treated as the basic unknown quantity. The search for the optimal ground state density has been simplified by a unitary transformation on the trial density which has been optimized by GA.